Concanate data from one table to another - sql

Question

I want to concanate the data from TABLE1 to TABLE 2
TABLE 1
id grp_name
-----------------------------
1 A#erf,R#erf.in
2 B#go.in,D#st.org/S#rec.uy
3 C#st.org,X#we.in,S#erl.in
4 D#gh.ou#F#rt.ot
5 E#rth.or
TABLE 2
code name
-----------------------------------
1 A#we.ot,D#ref.as
2 B#de.in
3 C#gr.cpm
4 D#yahoo.com
5 E#erf.com
6 F#google.com
I want to join grp_name data with name data like concanate using comma (',')
grp_name data having unwanted symboles like '#', '/', I want to elimate those too.
I created below procedure, but i dont know i effective or not.
If it is possible with simple update statement alone or merge statement alone let me know.
Excepted result
code name
1 A#we.ot,D#ref.as,A#erf,R#erf.in
2 B#de.in,B#go.in,D#st.org,S#rec.uy
3 C#gr.cpm,C#st.org,X#we.in,S#erl.in
4 D#yahoo.com,D#gh.ou,F#rt.ot
5 E#erf.com,E#rth.or
6 F#google.com
CREATE OR REPLACE PROCEDURE procedure1
AS
CURSOR cur
IS
SELECT id, grp_name
FROM TABLE 1;
CURSOR cur2
IS
SELECT code, name
FROM TABLE 2;
v_a VARCHAR2(300);
v_b VARCHAR2(25);
v_c VARCHAR2(4000);
v_d VARCHAR2(250);
BEGIN
FOR i IN cur
LOOP
v_a := ','||i.grp_name;
v_b := i.id;
FOR e IN cur2
LOOP
v_c := e.name || v_a ;
v_d := i.code;
UPDATE schema_name.TABLE 2
SET name = v_c
WHERE v_d = v_b;
END LOOP;
END LOOP;
-- COMMIT;
END;

Answer 1

In the simplest case, like #jarlh said, you can just do an update:
update table2
set name = name
|| (select ',' || regexp_replace(grp_name, '[/#]', ',') -- replace / and # with ,
from table1
where table1.id = table2.code)
where code in (select id from table1); -- only update matching rows
The last line is not strictly necessary, but it's good to avoid unnecessary updates.
If you also have rows in TABLE1 which don't have a match in TABLE2 that you want to add, use an insert:
insert into table2 (code, name)
select id, regexp_replace(grp_name, '[/#]', ',')
from table1
where id not in (select code from table2);
Edit: and like Gordon mentioned, you would have an easier time with the SQL if you stored your data like this:
code name
1 A#we.ot
1 D#ref.as
1 A#erf
1 R#erf.in
2 B#de.in
2 B#go.in
2 D#st.org
2 S#rec.uy
3 C#gr.cpm
3 C#st.org
3 X#we.in
3 S#erl.in

Answer 2

Please try something like this:
With tab1 as
(
select 1 as id,'A#er#f,R#erf.in' as grp_name from dual union all
select 2 as id,'B#go.in,D#st.or#g/S#rec.uy' as grp_name from dual
),
tab2 as (
select 1 as id,'A#we.ot,D#ref.as' as name from dual union all
select 2 as id,'B#de.in' as name from dual
)
select REGEXP_REPLACE (grp_name||','||name,'[^a-zA-Z0-9|\#|\,]','') as name from tab1 inner join tab2 using(id);
Result:
A#erf,R#erfin,A#weot,D#refas
B#goin,D#storgS#recuy,B#dein